This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [gomp] i386 cmpxchg8b and x86_64 cmpxchg16b support
- From: Richard Henderson <rth at redhat dot com>
- To: Andi Kleen <ak at suse dot de>
- Cc: Jakub Jelinek <jakub at redhat dot com>, gcc-patches at gcc dot gnu dot org
- Date: Mon, 10 Oct 2005 18:35:23 -0700
- Subject: Re: [gomp] i386 cmpxchg8b and x86_64 cmpxchg16b support
- References: <20051010141153.GQ22818@devserv.devel.redhat.com.suse.lists.egcs-patches> <p73psqduet1.fsf@verdi.suse.de> <20051011011229.GA6139@redhat.com> <200510110322.50310.ak@suse.de>
On Tue, Oct 11, 2005 at 03:22:49AM +0200, Andi Kleen wrote:
> 16 byte CMPXCHG16 is legal in 32bit code so I would have expected
> -march=nocona to emit it there too. If not it should.
Um, how? It gets the old and new values from dx:ax and bx:cx.
There's only 8 bytes in these paris in 32-bit code.
r~